| Date: | Sat, 24 Apr 1999 11:31:14 +0200 |
|---|
| Reply-To: | Albert Bonillo <abonillo@METODO.UAB.ES> |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU> |
|---|
| From: | Albert Bonillo <abonillo@METODO.UAB.ES> |
|---|
| Subject: | RV: IC 95% of the median in KM procedure |
|---|
| Content-Type: | multipart/alternative;
|
|---|
I made a Kaplan-Meier estimation survival's time. When I compared the ouput generated for SPSS 8.0 and BMDP PC90 I could see the great difference in the estimation of the 95% confidence interval of median: with the same data file, BMDP estimate the confidence interval about:
MEDIAN (50TH) 218.00 114.86
MEDIAN SURVIVAL TIME (85.00 , 392.00)
and SPSS ouput
Median: 218 115 ( 0, 443 ).
The rest of the ouput is equal in the two statistical packages.
What causes these differences?. How SPSS estimates the confidence interval of the median in the KM procedure?. I have read the Advanced Statistics Handbook of SPSS and there is not information about this estimation.
-----------------------------------------------------
Albert Bonillo Martín
abonilloX@metodo.uab.es
Universitat Auṭnoma de Barcelona
-----------------------------------------------------
Remove the "X" in the email adress
[text/html]
| 
